How to measure Insersion Loss of an End Fed Half Wave Transformer

In this video I show you how to measure insertion loss / efficiency of an End Fed Half Wave Transformer. Apologies for the contrast on the NanoVNA not being able to see the screen more clearly.
Let me know if you have any questions
#hamr #hamradio #ferrite #nanovna

[@KO4VNX]

I keep worrying about wire gauge for my antennas, but then I remembered no, it is the transformer that I need to worry about over-powering, so now I'm back to rewatching all of your EFHW videos all over again!

[@MM0OPXFieldRadio]

Theres a few. Probably a few more videos in me yet.

[@Htfhjooffhgghwwwwdf]

During the last few months I have tested lots of wire antennas. Forgetting the theory and looking at what works in practice, the EFHW was the shining light for multi band operation. I like to get out and activate parks under the World Wide Flora and Fauna program. (Also POTA) In a three hour session I normally have around 50 contacts including stations in New Zealand, Europe, Japan and the US as well as my home land in Australia of course. I build my own transformers and run 20W. Gotta love the EFHW.

[@richardholmes9469]

Kudos for going to the trouble of testing - and I enjoyed your video. Thank you. May I suggest calibrating at the fixture - screw terminals - where you connect the transformers as there will be losses in the BNC and the screw terminals. It would make an interesting follow up - and I may well be wrong. Yes it might not be a lot but it will add to uncertainty. I've used this method also and found it agrees with a simple link and then using Rp to calculate efficiency for the lower bands, but diverges from the single transformer and terminating resistor approach. From the shape of the S21 it is suggesting that an extra primary turn may help on 160m but will most likely make 10m worse. Getting broadband over 160m - 10m has defeated me in the past. You can do a heating calc from those figures also. Again, thanks for the video.

[@stevedick5479]

You can do a better test with the VNA. Use a single transformer, and for the output load, use a 2400 ohm resistor in series with a 52 ohm resistor. Feed the junction of those into the VNA which will see close to 50 ohms impedance. You have to correct for the attenuation due to voltage divider of the two resistors. This measurement eliminates the double hit of the two series leakage inductances of the two transformers. With the single transformer method, you only see effects of leakage inductance from the single transformer. -Steve K1RF

[@richardholmes9469]

I've done both. Calibration of the test fixture at the plane of the transformer is critical tho. Despite my best efforts I do get significant differences and I suspect my calibration is at fault. I've found the single transformer and terminating resistor method showed significantly better efficiency than the 2 transformer method - but I don't know which is correct. Bearing in mind that no antenna will ever present such a load - but I do like to be able to to compare transformers.

[@MM0OPXFieldRadio]

Hi Steve, I'm having trouble visualising what your describing. My background is not electronics. Can you direct me to a description of the method with pictures so I can try? I see you mentioned the losses above 15m can't be correct, can you quantify this through either published data or your own experimentation? I'm going to continue with the back to back for now as all my checks are then relative to each other. This is a huge learning journey for me.

[@evil_lair_electronics]

@@MM0OPXFieldRadio Over and over some people want to claim that the losses at the higher frequencies of HF cannot possibly be correct for this 3 stack of 52 mix. Yet somehow they seem to completely ignore that there are multiple 43 mix options which do not show this loss when doing exactly the same testing. If what they claim is true, then the 43 mix examples must also be showing false loss during testing and therefore also be even better than the measurements show. Now I certainly don't believe that and I doubt you do either.

[@Htfhjooffhgghwwwwdf]

That was very interesting as I use the 40m EFHW almost exclusively. I use 43 material in my toroids so not sure how much the results would differ between the two materials.

[@richardholmes9469]

In my experience, for the lower bands adding an extra turn to the primary makes a significant improvement, but can make 10m band worse. I use #43 and similar materials. Using a close wound auto transformer winding method improves performance by a few percent.

[@KK6USYHamRadioAdventures]

Great work Colin, very interesting to see the results!

[@MM0OPXFieldRadio]

It's an eye opener Chuck. This isn't one for 10m.

[@owlcricker-k7ulm]

It would be interesting to see if different mixes in the cores would improve the high bands of changes in capacitor values.

[@MM0OPXFieldRadio]

The combinations are endless to try. Theres maybe something in that.

[@richardholmes9469]

@@MM0OPXFieldRadio Indeed it is endless. I've spent 12 months trying many combinations in many ways. It is great to see someone actually do the testing. On the capacitor - I've been meaning to put a variable on the transformer and see what that does. Core geometry and winding technique can make big differences.

[@brendanhynes473]

Thanks Colin, nice one

[@MM0OPXFieldRadio]

Cheers Brendan 👍

[@mark_b]

Newbie question here: Is it possible to measure the loss and performance of a transformer purchased - without having a second identical one? I bought two to experiment with - but would love to know the characteristics of each.

[@MM0OPXFieldRadio]

Short answer is yes. I have a method someone gave me. I'm yet to try this but when i do i will make a video on it.

[@johnblake4503]

Really interesting video Colin, I do have a VNA and have experimented like this. I was also wondering, as I have a 2 ch Rigol scope and a signal generator whether another method could be to connect my generator (which has 50 ohm output impedance) to the transformer primary then compare the output level on the secondary. I suppose the only difference would be the scope has an high input impedance. Then I thought about putting a T piece on the secondary scope probe with a variable resistor and I could then adjust it for maximum output, measuring the resistance could then give me an accurate turns ration 2450/50 etc. Does any of this make any sense to you. Main advantage is I only need one ferrite to test.

[@MM0OPXFieldRadio]

It kinda does. Im no expert on the method. There is a method to use a 2450 non inductive resistor. I cant remember the method exactly but you only need one transformer. I need to look it up again.

[@ralph9987]

Colln, did you measure the loss of the two test leads incl connectors?

[@MM0OPXFieldRadio]

Yes, fitst thing I did after calibration. There was a 0.01db of loss across the total setup.

[@stevenmcc6052]

Good stuff. How much coupling did you get from having them next to each other? Is there an effect from the desk? Ie metal frame etc

[@MM0OPXFieldRadio]

I cant quantity that. My desk is 100% non conductive. Its a ply top sitting on top of plastic saw horses.

[@joeb3300]

I did notice a change in the frequency response as the transformer assembly was slid away from the VNA. I may try a similar experiment with the transformers supported away from the table-top (perhaps resting on a thin cardboard box).

[@danielhorvat4340]

So, it is evident that the loss of high bands has been confirmed so many times in the last 7 years. Unfortunately, amateurs keep looking only at SWR which is great on high bands due to simple attenuation of reflection on the way back thru the transformer, just like you see low SWR at the radio side when inserting long and lossy coax. The Insertion loss of less than 0.5 dB 3-30MHz is possible, see the MEF-330-1K+ with three 2.4" cores stacked.

[@MM0OPXFieldRadio]

100% agree, myantennas is the only manufacturer that publish insertion loss data.

[@wojteksz.8984]

H.A. 2 STACKED FT240-43 ?

[@MM0OPXFieldRadio]

No, still to check the 2 stacks

[@evil_lair_electronics]

Proven yet again that the 3 stack of 52 mix is NOT an efficient 80m-10m antenna. I can think of at least 4 people now that have shown this sort of measurement, but the masses keep building these inefficient transformers anyway. 🙄

[@Htfhjooffhgghwwwwdf]

What’s the alternative?

[@richardholmes9469]

I'd be very interested to see how a 3 turn primary works, and close wind the transformer. Agreed, the 2/14 and spread winding keeps on happening. But huge kudos to the people who actually do testing. Even more kudos to those who experiment and take it further.

[@MM0OPXFieldRadio]

Yep, there's unlimited options to try. I'm pretty sure doing what you suggest will improve the lower bands at the expense of the higher bands. This is what happened on the FT140-43 which has comparative geometry.

[@MrTalaue]

It won´t matter that much with a generally unefficient antenna.

[@MM0OPXFieldRadio]

How is it a generaly inefficient antenna. Its a dipole, yes mounting some ways it can be inefficient but mounted at the right height/configuration it's as good as it gets without adding reflectors/directors.

[@benitocarrilesgarcia9673]

Tu calibración no es correcta, tienes que calibrar con los dos cables juntos en serie, saludos.

[@MM0OPXFieldRadio]

Can you elaborate further? Yes I add on the BNC connectors etc after calibration but it makes no measurable difference at HF.